{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 47974,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002084462417142619,
      "grad_norm": 17.418737411499023,
      "learning_rate": 5.2111831991453666e-08,
      "loss": 6.7115,
      "step": 100
    },
    {
      "epoch": 0.004168924834285238,
      "grad_norm": 14.615851402282715,
      "learning_rate": 1.0422366398290733e-07,
      "loss": 6.6983,
      "step": 200
    },
    {
      "epoch": 0.006253387251427857,
      "grad_norm": 12.82895278930664,
      "learning_rate": 1.5633549597436098e-07,
      "loss": 6.6196,
      "step": 300
    },
    {
      "epoch": 0.008337849668570476,
      "grad_norm": 9.737526893615723,
      "learning_rate": 2.0844732796581466e-07,
      "loss": 6.4232,
      "step": 400
    },
    {
      "epoch": 0.010422312085713094,
      "grad_norm": 10.140837669372559,
      "learning_rate": 2.605591599572683e-07,
      "loss": 6.0683,
      "step": 500
    },
    {
      "epoch": 0.012506774502855714,
      "grad_norm": 6.8095598220825195,
      "learning_rate": 3.1267099194872195e-07,
      "loss": 5.6784,
      "step": 600
    },
    {
      "epoch": 0.014591236919998333,
      "grad_norm": 4.862451076507568,
      "learning_rate": 3.6478282394017564e-07,
      "loss": 5.3257,
      "step": 700
    },
    {
      "epoch": 0.016675699337140953,
      "grad_norm": 4.783125877380371,
      "learning_rate": 4.168946559316293e-07,
      "loss": 5.0905,
      "step": 800
    },
    {
      "epoch": 0.01876016175428357,
      "grad_norm": 2.310032367706299,
      "learning_rate": 4.6900648792308296e-07,
      "loss": 4.8735,
      "step": 900
    },
    {
      "epoch": 0.02084462417142619,
      "grad_norm": 1.856097936630249,
      "learning_rate": 5.211183199145366e-07,
      "loss": 4.7671,
      "step": 1000
    },
    {
      "epoch": 0.02292908658856881,
      "grad_norm": 1.9036556482315063,
      "learning_rate": 5.732301519059903e-07,
      "loss": 4.6835,
      "step": 1100
    },
    {
      "epoch": 0.025013549005711427,
      "grad_norm": 2.0767664909362793,
      "learning_rate": 6.253419838974439e-07,
      "loss": 4.5978,
      "step": 1200
    },
    {
      "epoch": 0.027098011422854045,
      "grad_norm": 5.158967018127441,
      "learning_rate": 6.774538158888977e-07,
      "loss": 4.5468,
      "step": 1300
    },
    {
      "epoch": 0.029182473839996666,
      "grad_norm": 2.4854819774627686,
      "learning_rate": 7.295656478803513e-07,
      "loss": 4.4797,
      "step": 1400
    },
    {
      "epoch": 0.031266936257139284,
      "grad_norm": 2.1669373512268066,
      "learning_rate": 7.81677479871805e-07,
      "loss": 4.4619,
      "step": 1500
    },
    {
      "epoch": 0.033351398674281905,
      "grad_norm": 3.4315173625946045,
      "learning_rate": 8.337893118632586e-07,
      "loss": 4.4004,
      "step": 1600
    },
    {
      "epoch": 0.03543586109142452,
      "grad_norm": 4.037099361419678,
      "learning_rate": 8.859011438547123e-07,
      "loss": 4.3784,
      "step": 1700
    },
    {
      "epoch": 0.03752032350856714,
      "grad_norm": 1.5988932847976685,
      "learning_rate": 9.380129758461659e-07,
      "loss": 4.3,
      "step": 1800
    },
    {
      "epoch": 0.03960478592570976,
      "grad_norm": 3.557048797607422,
      "learning_rate": 9.901248078376196e-07,
      "loss": 4.2595,
      "step": 1900
    },
    {
      "epoch": 0.04168924834285238,
      "grad_norm": 2.08551025390625,
      "learning_rate": 1.0422366398290733e-06,
      "loss": 4.2167,
      "step": 2000
    },
    {
      "epoch": 0.043773710759995,
      "grad_norm": 2.8505704402923584,
      "learning_rate": 1.094348471820527e-06,
      "loss": 4.177,
      "step": 2100
    },
    {
      "epoch": 0.04585817317713762,
      "grad_norm": 3.39355206489563,
      "learning_rate": 1.1464603038119807e-06,
      "loss": 4.1004,
      "step": 2200
    },
    {
      "epoch": 0.04794263559428023,
      "grad_norm": 4.5188093185424805,
      "learning_rate": 1.1985721358034343e-06,
      "loss": 3.9719,
      "step": 2300
    },
    {
      "epoch": 0.050027098011422855,
      "grad_norm": 6.547112464904785,
      "learning_rate": 1.2506839677948878e-06,
      "loss": 3.8793,
      "step": 2400
    },
    {
      "epoch": 0.052111560428565476,
      "grad_norm": 3.9441330432891846,
      "learning_rate": 1.3027957997863417e-06,
      "loss": 3.7997,
      "step": 2500
    },
    {
      "epoch": 0.05419602284570809,
      "grad_norm": 18.23339080810547,
      "learning_rate": 1.3549076317777954e-06,
      "loss": 3.6307,
      "step": 2600
    },
    {
      "epoch": 0.05628048526285071,
      "grad_norm": 3.295428991317749,
      "learning_rate": 1.4070194637692489e-06,
      "loss": 3.5646,
      "step": 2700
    },
    {
      "epoch": 0.05836494767999333,
      "grad_norm": 15.891700744628906,
      "learning_rate": 1.4591312957607026e-06,
      "loss": 3.4813,
      "step": 2800
    },
    {
      "epoch": 0.06044941009713595,
      "grad_norm": 5.232209205627441,
      "learning_rate": 1.5112431277521565e-06,
      "loss": 3.389,
      "step": 2900
    },
    {
      "epoch": 0.06253387251427857,
      "grad_norm": 9.18393611907959,
      "learning_rate": 1.56335495974361e-06,
      "loss": 3.3027,
      "step": 3000
    },
    {
      "epoch": 0.06461833493142119,
      "grad_norm": 12.729754447937012,
      "learning_rate": 1.6154667917350636e-06,
      "loss": 3.2259,
      "step": 3100
    },
    {
      "epoch": 0.06670279734856381,
      "grad_norm": 4.382412433624268,
      "learning_rate": 1.6675786237265173e-06,
      "loss": 3.1394,
      "step": 3200
    },
    {
      "epoch": 0.06878725976570642,
      "grad_norm": 5.850008964538574,
      "learning_rate": 1.7196904557179708e-06,
      "loss": 3.093,
      "step": 3300
    },
    {
      "epoch": 0.07087172218284904,
      "grad_norm": 3.021024227142334,
      "learning_rate": 1.7718022877094247e-06,
      "loss": 2.9828,
      "step": 3400
    },
    {
      "epoch": 0.07295618459999166,
      "grad_norm": 2.3273675441741943,
      "learning_rate": 1.8239141197008784e-06,
      "loss": 2.921,
      "step": 3500
    },
    {
      "epoch": 0.07504064701713428,
      "grad_norm": 3.740726947784424,
      "learning_rate": 1.8760259516923318e-06,
      "loss": 2.8224,
      "step": 3600
    },
    {
      "epoch": 0.0771251094342769,
      "grad_norm": 2.0490314960479736,
      "learning_rate": 1.9281377836837857e-06,
      "loss": 2.8053,
      "step": 3700
    },
    {
      "epoch": 0.07920957185141952,
      "grad_norm": 4.149216651916504,
      "learning_rate": 1.980249615675239e-06,
      "loss": 2.8044,
      "step": 3800
    },
    {
      "epoch": 0.08129403426856213,
      "grad_norm": 2.3400914669036865,
      "learning_rate": 2.0323614476666927e-06,
      "loss": 2.7467,
      "step": 3900
    },
    {
      "epoch": 0.08337849668570475,
      "grad_norm": 8.31890869140625,
      "learning_rate": 2.0844732796581466e-06,
      "loss": 2.7686,
      "step": 4000
    },
    {
      "epoch": 0.08546295910284737,
      "grad_norm": 5.363232135772705,
      "learning_rate": 2.1365851116496005e-06,
      "loss": 2.7288,
      "step": 4100
    },
    {
      "epoch": 0.08754742151999,
      "grad_norm": 2.7400481700897217,
      "learning_rate": 2.188696943641054e-06,
      "loss": 2.6961,
      "step": 4200
    },
    {
      "epoch": 0.08963188393713262,
      "grad_norm": 1.5836577415466309,
      "learning_rate": 2.2408087756325074e-06,
      "loss": 2.6586,
      "step": 4300
    },
    {
      "epoch": 0.09171634635427524,
      "grad_norm": 1.6861375570297241,
      "learning_rate": 2.2929206076239613e-06,
      "loss": 2.6471,
      "step": 4400
    },
    {
      "epoch": 0.09380080877141785,
      "grad_norm": 2.1874806880950928,
      "learning_rate": 2.3450324396154148e-06,
      "loss": 2.64,
      "step": 4500
    },
    {
      "epoch": 0.09588527118856047,
      "grad_norm": 1.5436290502548218,
      "learning_rate": 2.3971442716068687e-06,
      "loss": 2.6245,
      "step": 4600
    },
    {
      "epoch": 0.09796973360570309,
      "grad_norm": 1.6910173892974854,
      "learning_rate": 2.449256103598322e-06,
      "loss": 2.5595,
      "step": 4700
    },
    {
      "epoch": 0.10005419602284571,
      "grad_norm": 1.635819673538208,
      "learning_rate": 2.5013679355897756e-06,
      "loss": 2.5527,
      "step": 4800
    },
    {
      "epoch": 0.10213865843998833,
      "grad_norm": 1.5254827737808228,
      "learning_rate": 2.5534797675812295e-06,
      "loss": 2.5624,
      "step": 4900
    },
    {
      "epoch": 0.10422312085713095,
      "grad_norm": 2.539045810699463,
      "learning_rate": 2.6055915995726834e-06,
      "loss": 2.5594,
      "step": 5000
    },
    {
      "epoch": 0.10630758327427356,
      "grad_norm": 1.8983627557754517,
      "learning_rate": 2.657703431564137e-06,
      "loss": 2.5547,
      "step": 5100
    },
    {
      "epoch": 0.10839204569141618,
      "grad_norm": 1.4213758707046509,
      "learning_rate": 2.709815263555591e-06,
      "loss": 2.5026,
      "step": 5200
    },
    {
      "epoch": 0.1104765081085588,
      "grad_norm": 1.7480638027191162,
      "learning_rate": 2.761927095547044e-06,
      "loss": 2.514,
      "step": 5300
    },
    {
      "epoch": 0.11256097052570142,
      "grad_norm": 1.9647337198257446,
      "learning_rate": 2.8140389275384977e-06,
      "loss": 2.5433,
      "step": 5400
    },
    {
      "epoch": 0.11464543294284404,
      "grad_norm": 2.1344046592712402,
      "learning_rate": 2.8661507595299516e-06,
      "loss": 2.5185,
      "step": 5500
    },
    {
      "epoch": 0.11672989535998667,
      "grad_norm": 2.501701593399048,
      "learning_rate": 2.918262591521405e-06,
      "loss": 2.516,
      "step": 5600
    },
    {
      "epoch": 0.11881435777712927,
      "grad_norm": 3.260753631591797,
      "learning_rate": 2.970374423512859e-06,
      "loss": 2.4857,
      "step": 5700
    },
    {
      "epoch": 0.1208988201942719,
      "grad_norm": 2.426595687866211,
      "learning_rate": 3.022486255504313e-06,
      "loss": 2.4193,
      "step": 5800
    },
    {
      "epoch": 0.12298328261141452,
      "grad_norm": 2.5531206130981445,
      "learning_rate": 3.074598087495766e-06,
      "loss": 2.458,
      "step": 5900
    },
    {
      "epoch": 0.12506774502855714,
      "grad_norm": 4.4168243408203125,
      "learning_rate": 3.12670991948722e-06,
      "loss": 2.4794,
      "step": 6000
    },
    {
      "epoch": 0.12715220744569974,
      "grad_norm": 3.948293924331665,
      "learning_rate": 3.1788217514786733e-06,
      "loss": 2.4691,
      "step": 6100
    },
    {
      "epoch": 0.12923666986284238,
      "grad_norm": 1.663045048713684,
      "learning_rate": 3.2309335834701272e-06,
      "loss": 2.4494,
      "step": 6200
    },
    {
      "epoch": 0.131321132279985,
      "grad_norm": 3.5404489040374756,
      "learning_rate": 3.283045415461581e-06,
      "loss": 2.428,
      "step": 6300
    },
    {
      "epoch": 0.13340559469712762,
      "grad_norm": 1.4624788761138916,
      "learning_rate": 3.3351572474530346e-06,
      "loss": 2.3999,
      "step": 6400
    },
    {
      "epoch": 0.13549005711427023,
      "grad_norm": 2.683394432067871,
      "learning_rate": 3.387269079444488e-06,
      "loss": 2.417,
      "step": 6500
    },
    {
      "epoch": 0.13757451953141284,
      "grad_norm": 2.5286734104156494,
      "learning_rate": 3.4393809114359415e-06,
      "loss": 2.423,
      "step": 6600
    },
    {
      "epoch": 0.13965898194855547,
      "grad_norm": 1.595863938331604,
      "learning_rate": 3.4914927434273954e-06,
      "loss": 2.433,
      "step": 6700
    },
    {
      "epoch": 0.14174344436569808,
      "grad_norm": 5.466402053833008,
      "learning_rate": 3.5436045754188493e-06,
      "loss": 2.4116,
      "step": 6800
    },
    {
      "epoch": 0.14382790678284071,
      "grad_norm": 1.5954893827438354,
      "learning_rate": 3.595716407410303e-06,
      "loss": 2.3939,
      "step": 6900
    },
    {
      "epoch": 0.14591236919998332,
      "grad_norm": 2.247056007385254,
      "learning_rate": 3.6478282394017567e-06,
      "loss": 2.3678,
      "step": 7000
    },
    {
      "epoch": 0.14799683161712596,
      "grad_norm": 5.489955425262451,
      "learning_rate": 3.6999400713932098e-06,
      "loss": 2.3802,
      "step": 7100
    },
    {
      "epoch": 0.15008129403426856,
      "grad_norm": 1.5120584964752197,
      "learning_rate": 3.7520519033846637e-06,
      "loss": 2.3891,
      "step": 7200
    },
    {
      "epoch": 0.15216575645141117,
      "grad_norm": 2.228731870651245,
      "learning_rate": 3.8041637353761176e-06,
      "loss": 2.3644,
      "step": 7300
    },
    {
      "epoch": 0.1542502188685538,
      "grad_norm": 1.7720115184783936,
      "learning_rate": 3.8562755673675715e-06,
      "loss": 2.3507,
      "step": 7400
    },
    {
      "epoch": 0.15633468128569641,
      "grad_norm": 4.0883331298828125,
      "learning_rate": 3.9083873993590245e-06,
      "loss": 2.3572,
      "step": 7500
    },
    {
      "epoch": 0.15841914370283905,
      "grad_norm": 1.3633663654327393,
      "learning_rate": 3.960499231350478e-06,
      "loss": 2.3387,
      "step": 7600
    },
    {
      "epoch": 0.16050360611998166,
      "grad_norm": 3.6679422855377197,
      "learning_rate": 4.0126110633419315e-06,
      "loss": 2.3446,
      "step": 7700
    },
    {
      "epoch": 0.16258806853712426,
      "grad_norm": 2.032133102416992,
      "learning_rate": 4.064722895333385e-06,
      "loss": 2.3273,
      "step": 7800
    },
    {
      "epoch": 0.1646725309542669,
      "grad_norm": 1.4509003162384033,
      "learning_rate": 4.116834727324839e-06,
      "loss": 2.3072,
      "step": 7900
    },
    {
      "epoch": 0.1667569933714095,
      "grad_norm": 4.3479413986206055,
      "learning_rate": 4.168946559316293e-06,
      "loss": 2.3016,
      "step": 8000
    },
    {
      "epoch": 0.16884145578855214,
      "grad_norm": 12.375444412231445,
      "learning_rate": 4.221058391307747e-06,
      "loss": 2.3145,
      "step": 8100
    },
    {
      "epoch": 0.17092591820569475,
      "grad_norm": 1.4382028579711914,
      "learning_rate": 4.273170223299201e-06,
      "loss": 2.274,
      "step": 8200
    },
    {
      "epoch": 0.17301038062283736,
      "grad_norm": 7.938540458679199,
      "learning_rate": 4.325282055290654e-06,
      "loss": 2.2873,
      "step": 8300
    },
    {
      "epoch": 0.17509484303998,
      "grad_norm": 2.1342923641204834,
      "learning_rate": 4.377393887282108e-06,
      "loss": 2.2906,
      "step": 8400
    },
    {
      "epoch": 0.1771793054571226,
      "grad_norm": 1.8758960962295532,
      "learning_rate": 4.429505719273561e-06,
      "loss": 2.2879,
      "step": 8500
    },
    {
      "epoch": 0.17926376787426523,
      "grad_norm": 6.642971038818359,
      "learning_rate": 4.481617551265015e-06,
      "loss": 2.2867,
      "step": 8600
    },
    {
      "epoch": 0.18134823029140784,
      "grad_norm": 5.171383380889893,
      "learning_rate": 4.533729383256469e-06,
      "loss": 2.2571,
      "step": 8700
    },
    {
      "epoch": 0.18343269270855048,
      "grad_norm": 1.664364218711853,
      "learning_rate": 4.585841215247923e-06,
      "loss": 2.2709,
      "step": 8800
    },
    {
      "epoch": 0.18551715512569308,
      "grad_norm": 7.0319719314575195,
      "learning_rate": 4.6379530472393765e-06,
      "loss": 2.2643,
      "step": 8900
    },
    {
      "epoch": 0.1876016175428357,
      "grad_norm": 2.7922122478485107,
      "learning_rate": 4.6900648792308296e-06,
      "loss": 2.2473,
      "step": 9000
    },
    {
      "epoch": 0.18968607995997833,
      "grad_norm": 1.8431055545806885,
      "learning_rate": 4.7421767112222835e-06,
      "loss": 2.2499,
      "step": 9100
    },
    {
      "epoch": 0.19177054237712093,
      "grad_norm": 3.7758209705352783,
      "learning_rate": 4.794288543213737e-06,
      "loss": 2.2595,
      "step": 9200
    },
    {
      "epoch": 0.19385500479426357,
      "grad_norm": 2.5738108158111572,
      "learning_rate": 4.84640037520519e-06,
      "loss": 2.2364,
      "step": 9300
    },
    {
      "epoch": 0.19593946721140618,
      "grad_norm": 9.412129402160645,
      "learning_rate": 4.898512207196644e-06,
      "loss": 2.2371,
      "step": 9400
    },
    {
      "epoch": 0.19802392962854878,
      "grad_norm": 1.357433795928955,
      "learning_rate": 4.950624039188098e-06,
      "loss": 2.2418,
      "step": 9500
    },
    {
      "epoch": 0.20010839204569142,
      "grad_norm": 9.067081451416016,
      "learning_rate": 5.002735871179551e-06,
      "loss": 2.2445,
      "step": 9600
    },
    {
      "epoch": 0.20219285446283403,
      "grad_norm": 2.770761728286743,
      "learning_rate": 5.054847703171005e-06,
      "loss": 2.2224,
      "step": 9700
    },
    {
      "epoch": 0.20427731687997666,
      "grad_norm": 1.7720422744750977,
      "learning_rate": 5.106959535162459e-06,
      "loss": 2.2507,
      "step": 9800
    },
    {
      "epoch": 0.20636177929711927,
      "grad_norm": 5.686391830444336,
      "learning_rate": 5.159071367153913e-06,
      "loss": 2.233,
      "step": 9900
    },
    {
      "epoch": 0.2084462417142619,
      "grad_norm": 6.246903419494629,
      "learning_rate": 5.211183199145367e-06,
      "loss": 2.2266,
      "step": 10000
    },
    {
      "epoch": 0.2105307041314045,
      "grad_norm": 7.122190475463867,
      "learning_rate": 5.26329503113682e-06,
      "loss": 2.239,
      "step": 10100
    },
    {
      "epoch": 0.21261516654854712,
      "grad_norm": 2.073593854904175,
      "learning_rate": 5.315406863128274e-06,
      "loss": 2.2405,
      "step": 10200
    },
    {
      "epoch": 0.21469962896568975,
      "grad_norm": 2.734534740447998,
      "learning_rate": 5.367518695119728e-06,
      "loss": 2.2019,
      "step": 10300
    },
    {
      "epoch": 0.21678409138283236,
      "grad_norm": 3.9620697498321533,
      "learning_rate": 5.419630527111182e-06,
      "loss": 2.2132,
      "step": 10400
    },
    {
      "epoch": 0.218868553799975,
      "grad_norm": 4.667324542999268,
      "learning_rate": 5.471742359102634e-06,
      "loss": 2.2028,
      "step": 10500
    },
    {
      "epoch": 0.2209530162171176,
      "grad_norm": 2.702505111694336,
      "learning_rate": 5.523854191094088e-06,
      "loss": 2.2188,
      "step": 10600
    },
    {
      "epoch": 0.2230374786342602,
      "grad_norm": 2.8783977031707764,
      "learning_rate": 5.575966023085542e-06,
      "loss": 2.2058,
      "step": 10700
    },
    {
      "epoch": 0.22512194105140285,
      "grad_norm": 4.026909351348877,
      "learning_rate": 5.6280778550769955e-06,
      "loss": 2.2147,
      "step": 10800
    },
    {
      "epoch": 0.22720640346854545,
      "grad_norm": 2.730255126953125,
      "learning_rate": 5.680189687068449e-06,
      "loss": 2.203,
      "step": 10900
    },
    {
      "epoch": 0.2292908658856881,
      "grad_norm": 1.5225945711135864,
      "learning_rate": 5.732301519059903e-06,
      "loss": 2.1795,
      "step": 11000
    },
    {
      "epoch": 0.2313753283028307,
      "grad_norm": 2.083085775375366,
      "learning_rate": 5.784413351051356e-06,
      "loss": 2.1758,
      "step": 11100
    },
    {
      "epoch": 0.23345979071997333,
      "grad_norm": 3.0502681732177734,
      "learning_rate": 5.83652518304281e-06,
      "loss": 2.1901,
      "step": 11200
    },
    {
      "epoch": 0.23554425313711594,
      "grad_norm": 1.8443273305892944,
      "learning_rate": 5.888637015034264e-06,
      "loss": 2.1827,
      "step": 11300
    },
    {
      "epoch": 0.23762871555425855,
      "grad_norm": 2.201040506362915,
      "learning_rate": 5.940748847025718e-06,
      "loss": 2.1681,
      "step": 11400
    },
    {
      "epoch": 0.23971317797140118,
      "grad_norm": 4.19305419921875,
      "learning_rate": 5.992860679017172e-06,
      "loss": 2.1854,
      "step": 11500
    },
    {
      "epoch": 0.2417976403885438,
      "grad_norm": 1.3625741004943848,
      "learning_rate": 6.044972511008626e-06,
      "loss": 2.1777,
      "step": 11600
    },
    {
      "epoch": 0.24388210280568642,
      "grad_norm": 2.7626659870147705,
      "learning_rate": 6.097084343000078e-06,
      "loss": 2.1823,
      "step": 11700
    },
    {
      "epoch": 0.24596656522282903,
      "grad_norm": 5.543747901916504,
      "learning_rate": 6.149196174991532e-06,
      "loss": 2.1628,
      "step": 11800
    },
    {
      "epoch": 0.24805102763997164,
      "grad_norm": 1.4751501083374023,
      "learning_rate": 6.201308006982986e-06,
      "loss": 2.162,
      "step": 11900
    },
    {
      "epoch": 0.2501354900571143,
      "grad_norm": 2.3164498805999756,
      "learning_rate": 6.25341983897444e-06,
      "loss": 2.1571,
      "step": 12000
    },
    {
      "epoch": 0.2522199524742569,
      "grad_norm": 1.5216565132141113,
      "learning_rate": 6.305531670965893e-06,
      "loss": 2.1822,
      "step": 12100
    },
    {
      "epoch": 0.2543044148913995,
      "grad_norm": 2.282949924468994,
      "learning_rate": 6.357643502957347e-06,
      "loss": 2.1718,
      "step": 12200
    },
    {
      "epoch": 0.25638887730854215,
      "grad_norm": 1.349082112312317,
      "learning_rate": 6.4097553349488006e-06,
      "loss": 2.1843,
      "step": 12300
    },
    {
      "epoch": 0.25847333972568476,
      "grad_norm": 3.20103120803833,
      "learning_rate": 6.4618671669402545e-06,
      "loss": 2.1698,
      "step": 12400
    },
    {
      "epoch": 0.26055780214282737,
      "grad_norm": 1.9603382349014282,
      "learning_rate": 6.513978998931708e-06,
      "loss": 2.1502,
      "step": 12500
    },
    {
      "epoch": 0.26264226455997,
      "grad_norm": 3.995859384536743,
      "learning_rate": 6.566090830923162e-06,
      "loss": 2.1536,
      "step": 12600
    },
    {
      "epoch": 0.2647267269771126,
      "grad_norm": 4.198755264282227,
      "learning_rate": 6.618202662914615e-06,
      "loss": 2.1655,
      "step": 12700
    },
    {
      "epoch": 0.26681118939425524,
      "grad_norm": 9.24932861328125,
      "learning_rate": 6.670314494906069e-06,
      "loss": 2.1528,
      "step": 12800
    },
    {
      "epoch": 0.26889565181139785,
      "grad_norm": 2.8210318088531494,
      "learning_rate": 6.722426326897523e-06,
      "loss": 2.1388,
      "step": 12900
    },
    {
      "epoch": 0.27098011422854046,
      "grad_norm": 1.819380760192871,
      "learning_rate": 6.774538158888976e-06,
      "loss": 2.1477,
      "step": 13000
    },
    {
      "epoch": 0.27306457664568307,
      "grad_norm": 2.15307354927063,
      "learning_rate": 6.826649990880429e-06,
      "loss": 2.1549,
      "step": 13100
    },
    {
      "epoch": 0.2751490390628257,
      "grad_norm": 3.6344521045684814,
      "learning_rate": 6.878761822871883e-06,
      "loss": 2.1478,
      "step": 13200
    },
    {
      "epoch": 0.27723350147996834,
      "grad_norm": 3.5121541023254395,
      "learning_rate": 6.930873654863337e-06,
      "loss": 2.1366,
      "step": 13300
    },
    {
      "epoch": 0.27931796389711094,
      "grad_norm": 1.6062060594558716,
      "learning_rate": 6.982985486854791e-06,
      "loss": 2.1338,
      "step": 13400
    },
    {
      "epoch": 0.28140242631425355,
      "grad_norm": 3.3421785831451416,
      "learning_rate": 7.035097318846245e-06,
      "loss": 2.1485,
      "step": 13500
    },
    {
      "epoch": 0.28348688873139616,
      "grad_norm": 1.5233893394470215,
      "learning_rate": 7.087209150837699e-06,
      "loss": 2.1412,
      "step": 13600
    },
    {
      "epoch": 0.28557135114853877,
      "grad_norm": 12.393982887268066,
      "learning_rate": 7.139320982829152e-06,
      "loss": 2.1399,
      "step": 13700
    },
    {
      "epoch": 0.28765581356568143,
      "grad_norm": 2.9735476970672607,
      "learning_rate": 7.191432814820606e-06,
      "loss": 2.143,
      "step": 13800
    },
    {
      "epoch": 0.28974027598282404,
      "grad_norm": 2.875746965408325,
      "learning_rate": 7.2435446468120595e-06,
      "loss": 2.1396,
      "step": 13900
    },
    {
      "epoch": 0.29182473839996664,
      "grad_norm": 1.34048330783844,
      "learning_rate": 7.2956564788035134e-06,
      "loss": 2.1132,
      "step": 14000
    },
    {
      "epoch": 0.29390920081710925,
      "grad_norm": 1.4782146215438843,
      "learning_rate": 7.347768310794967e-06,
      "loss": 2.1326,
      "step": 14100
    },
    {
      "epoch": 0.2959936632342519,
      "grad_norm": 2.069708824157715,
      "learning_rate": 7.3998801427864195e-06,
      "loss": 2.1433,
      "step": 14200
    },
    {
      "epoch": 0.2980781256513945,
      "grad_norm": 1.770173192024231,
      "learning_rate": 7.451991974777873e-06,
      "loss": 2.1154,
      "step": 14300
    },
    {
      "epoch": 0.30016258806853713,
      "grad_norm": 2.3121659755706787,
      "learning_rate": 7.504103806769327e-06,
      "loss": 2.1279,
      "step": 14400
    },
    {
      "epoch": 0.30224705048567974,
      "grad_norm": 1.7393426895141602,
      "learning_rate": 7.556215638760781e-06,
      "loss": 2.1328,
      "step": 14500
    },
    {
      "epoch": 0.30433151290282234,
      "grad_norm": 1.6336350440979004,
      "learning_rate": 7.608327470752235e-06,
      "loss": 2.0962,
      "step": 14600
    },
    {
      "epoch": 0.306415975319965,
      "grad_norm": 1.4256808757781982,
      "learning_rate": 7.660439302743689e-06,
      "loss": 2.1283,
      "step": 14700
    },
    {
      "epoch": 0.3085004377371076,
      "grad_norm": 1.3236781358718872,
      "learning_rate": 7.712551134735143e-06,
      "loss": 2.1225,
      "step": 14800
    },
    {
      "epoch": 0.3105849001542502,
      "grad_norm": 15.130003929138184,
      "learning_rate": 7.764662966726597e-06,
      "loss": 2.1256,
      "step": 14900
    },
    {
      "epoch": 0.31266936257139283,
      "grad_norm": 3.1798925399780273,
      "learning_rate": 7.816774798718049e-06,
      "loss": 2.1199,
      "step": 15000
    },
    {
      "epoch": 0.31475382498853544,
      "grad_norm": 3.058833599090576,
      "learning_rate": 7.868886630709503e-06,
      "loss": 2.1067,
      "step": 15100
    },
    {
      "epoch": 0.3168382874056781,
      "grad_norm": 2.1483590602874756,
      "learning_rate": 7.920998462700957e-06,
      "loss": 2.0995,
      "step": 15200
    },
    {
      "epoch": 0.3189227498228207,
      "grad_norm": 1.3835676908493042,
      "learning_rate": 7.97311029469241e-06,
      "loss": 2.1013,
      "step": 15300
    },
    {
      "epoch": 0.3210072122399633,
      "grad_norm": 1.9250987768173218,
      "learning_rate": 8.025222126683863e-06,
      "loss": 2.1124,
      "step": 15400
    },
    {
      "epoch": 0.3230916746571059,
      "grad_norm": 10.668524742126465,
      "learning_rate": 8.077333958675317e-06,
      "loss": 2.0705,
      "step": 15500
    },
    {
      "epoch": 0.32517613707424853,
      "grad_norm": 1.4059637784957886,
      "learning_rate": 8.12944579066677e-06,
      "loss": 2.1161,
      "step": 15600
    },
    {
      "epoch": 0.3272605994913912,
      "grad_norm": 1.4063282012939453,
      "learning_rate": 8.181557622658225e-06,
      "loss": 2.1158,
      "step": 15700
    },
    {
      "epoch": 0.3293450619085338,
      "grad_norm": 1.4921411275863647,
      "learning_rate": 8.233669454649678e-06,
      "loss": 2.0935,
      "step": 15800
    },
    {
      "epoch": 0.3314295243256764,
      "grad_norm": 1.4278446435928345,
      "learning_rate": 8.285781286641132e-06,
      "loss": 2.0921,
      "step": 15900
    },
    {
      "epoch": 0.333513986742819,
      "grad_norm": 1.509752631187439,
      "learning_rate": 8.337893118632586e-06,
      "loss": 2.0946,
      "step": 16000
    },
    {
      "epoch": 0.3355984491599616,
      "grad_norm": 3.8793394565582275,
      "learning_rate": 8.39000495062404e-06,
      "loss": 2.0952,
      "step": 16100
    },
    {
      "epoch": 0.3376829115771043,
      "grad_norm": 2.2685413360595703,
      "learning_rate": 8.442116782615494e-06,
      "loss": 2.0802,
      "step": 16200
    },
    {
      "epoch": 0.3397673739942469,
      "grad_norm": 1.4082143306732178,
      "learning_rate": 8.494228614606948e-06,
      "loss": 2.0913,
      "step": 16300
    },
    {
      "epoch": 0.3418518364113895,
      "grad_norm": 2.058350086212158,
      "learning_rate": 8.546340446598402e-06,
      "loss": 2.1075,
      "step": 16400
    },
    {
      "epoch": 0.3439362988285321,
      "grad_norm": 1.5691351890563965,
      "learning_rate": 8.598452278589856e-06,
      "loss": 2.0984,
      "step": 16500
    },
    {
      "epoch": 0.3460207612456747,
      "grad_norm": 2.385117769241333,
      "learning_rate": 8.650564110581308e-06,
      "loss": 2.0793,
      "step": 16600
    },
    {
      "epoch": 0.3481052236628174,
      "grad_norm": 2.0215940475463867,
      "learning_rate": 8.702675942572762e-06,
      "loss": 2.0739,
      "step": 16700
    },
    {
      "epoch": 0.35018968607996,
      "grad_norm": 1.5603607892990112,
      "learning_rate": 8.754787774564216e-06,
      "loss": 2.0733,
      "step": 16800
    },
    {
      "epoch": 0.3522741484971026,
      "grad_norm": 1.732659101486206,
      "learning_rate": 8.80689960655567e-06,
      "loss": 2.0928,
      "step": 16900
    },
    {
      "epoch": 0.3543586109142452,
      "grad_norm": 18.339719772338867,
      "learning_rate": 8.859011438547122e-06,
      "loss": 2.0802,
      "step": 17000
    },
    {
      "epoch": 0.35644307333138786,
      "grad_norm": 1.269181251525879,
      "learning_rate": 8.911123270538576e-06,
      "loss": 2.063,
      "step": 17100
    },
    {
      "epoch": 0.35852753574853047,
      "grad_norm": 1.4720960855484009,
      "learning_rate": 8.96323510253003e-06,
      "loss": 2.0952,
      "step": 17200
    },
    {
      "epoch": 0.3606119981656731,
      "grad_norm": 1.8136069774627686,
      "learning_rate": 9.015346934521484e-06,
      "loss": 2.0834,
      "step": 17300
    },
    {
      "epoch": 0.3626964605828157,
      "grad_norm": 1.4650546312332153,
      "learning_rate": 9.067458766512937e-06,
      "loss": 2.0775,
      "step": 17400
    },
    {
      "epoch": 0.3647809229999583,
      "grad_norm": 1.538031816482544,
      "learning_rate": 9.119570598504391e-06,
      "loss": 2.094,
      "step": 17500
    },
    {
      "epoch": 0.36686538541710095,
      "grad_norm": 1.2462719678878784,
      "learning_rate": 9.171682430495845e-06,
      "loss": 2.072,
      "step": 17600
    },
    {
      "epoch": 0.36894984783424356,
      "grad_norm": 1.4545553922653198,
      "learning_rate": 9.223794262487299e-06,
      "loss": 2.0787,
      "step": 17700
    },
    {
      "epoch": 0.37103431025138617,
      "grad_norm": 1.3159983158111572,
      "learning_rate": 9.275906094478753e-06,
      "loss": 2.0681,
      "step": 17800
    },
    {
      "epoch": 0.3731187726685288,
      "grad_norm": 4.405596733093262,
      "learning_rate": 9.328017926470205e-06,
      "loss": 2.0666,
      "step": 17900
    },
    {
      "epoch": 0.3752032350856714,
      "grad_norm": 1.6247169971466064,
      "learning_rate": 9.380129758461659e-06,
      "loss": 2.0682,
      "step": 18000
    },
    {
      "epoch": 0.37728769750281405,
      "grad_norm": 1.2033179998397827,
      "learning_rate": 9.432241590453113e-06,
      "loss": 2.0714,
      "step": 18100
    },
    {
      "epoch": 0.37937215991995665,
      "grad_norm": 1.3838162422180176,
      "learning_rate": 9.484353422444567e-06,
      "loss": 2.0692,
      "step": 18200
    },
    {
      "epoch": 0.38145662233709926,
      "grad_norm": 1.642264723777771,
      "learning_rate": 9.53646525443602e-06,
      "loss": 2.0674,
      "step": 18300
    },
    {
      "epoch": 0.38354108475424187,
      "grad_norm": 1.152984619140625,
      "learning_rate": 9.588577086427475e-06,
      "loss": 2.0644,
      "step": 18400
    },
    {
      "epoch": 0.3856255471713845,
      "grad_norm": 3.9230525493621826,
      "learning_rate": 9.640688918418929e-06,
      "loss": 2.0642,
      "step": 18500
    },
    {
      "epoch": 0.38771000958852714,
      "grad_norm": 1.4267147779464722,
      "learning_rate": 9.69280075041038e-06,
      "loss": 2.0842,
      "step": 18600
    },
    {
      "epoch": 0.38979447200566975,
      "grad_norm": 2.8921425342559814,
      "learning_rate": 9.744912582401835e-06,
      "loss": 2.0604,
      "step": 18700
    },
    {
      "epoch": 0.39187893442281235,
      "grad_norm": 1.3720977306365967,
      "learning_rate": 9.797024414393289e-06,
      "loss": 2.0512,
      "step": 18800
    },
    {
      "epoch": 0.39396339683995496,
      "grad_norm": 2.229217052459717,
      "learning_rate": 9.849136246384743e-06,
      "loss": 2.0377,
      "step": 18900
    },
    {
      "epoch": 0.39604785925709757,
      "grad_norm": 1.2066770792007446,
      "learning_rate": 9.901248078376196e-06,
      "loss": 2.0627,
      "step": 19000
    },
    {
      "epoch": 0.39813232167424023,
      "grad_norm": 1.3302040100097656,
      "learning_rate": 9.953359910367649e-06,
      "loss": 2.063,
      "step": 19100
    },
    {
      "epoch": 0.40021678409138284,
      "grad_norm": 2.0192320346832275,
      "learning_rate": 1.0005471742359103e-05,
      "loss": 2.0521,
      "step": 19200
    },
    {
      "epoch": 0.40230124650852545,
      "grad_norm": 1.3492109775543213,
      "learning_rate": 1.0057583574350558e-05,
      "loss": 2.0433,
      "step": 19300
    },
    {
      "epoch": 0.40438570892566805,
      "grad_norm": 1.3807893991470337,
      "learning_rate": 1.010969540634201e-05,
      "loss": 2.0624,
      "step": 19400
    },
    {
      "epoch": 0.4064701713428107,
      "grad_norm": 1.525590419769287,
      "learning_rate": 1.0161807238333466e-05,
      "loss": 2.0475,
      "step": 19500
    },
    {
      "epoch": 0.4085546337599533,
      "grad_norm": 1.2582364082336426,
      "learning_rate": 1.0213919070324918e-05,
      "loss": 2.0339,
      "step": 19600
    },
    {
      "epoch": 0.41063909617709593,
      "grad_norm": 2.245859146118164,
      "learning_rate": 1.026603090231637e-05,
      "loss": 2.0401,
      "step": 19700
    },
    {
      "epoch": 0.41272355859423854,
      "grad_norm": 1.3547762632369995,
      "learning_rate": 1.0318142734307826e-05,
      "loss": 2.035,
      "step": 19800
    },
    {
      "epoch": 0.41480802101138115,
      "grad_norm": 1.5345648527145386,
      "learning_rate": 1.0370254566299278e-05,
      "loss": 2.048,
      "step": 19900
    },
    {
      "epoch": 0.4168924834285238,
      "grad_norm": 9.708149909973145,
      "learning_rate": 1.0422366398290734e-05,
      "loss": 2.0277,
      "step": 20000
    },
    {
      "epoch": 0.4189769458456664,
      "grad_norm": 1.277744174003601,
      "learning_rate": 1.0474478230282186e-05,
      "loss": 2.046,
      "step": 20100
    },
    {
      "epoch": 0.421061408262809,
      "grad_norm": 1.1756649017333984,
      "learning_rate": 1.052659006227364e-05,
      "loss": 2.0232,
      "step": 20200
    },
    {
      "epoch": 0.42314587067995163,
      "grad_norm": 1.0774590969085693,
      "learning_rate": 1.0578701894265094e-05,
      "loss": 2.0486,
      "step": 20300
    },
    {
      "epoch": 0.42523033309709424,
      "grad_norm": 1.1391072273254395,
      "learning_rate": 1.0630813726256548e-05,
      "loss": 2.0329,
      "step": 20400
    },
    {
      "epoch": 0.4273147955142369,
      "grad_norm": 1.9197932481765747,
      "learning_rate": 1.0682925558248001e-05,
      "loss": 2.0328,
      "step": 20500
    },
    {
      "epoch": 0.4293992579313795,
      "grad_norm": 1.3254817724227905,
      "learning_rate": 1.0735037390239455e-05,
      "loss": 2.0308,
      "step": 20600
    },
    {
      "epoch": 0.4314837203485221,
      "grad_norm": 4.676088809967041,
      "learning_rate": 1.0787149222230908e-05,
      "loss": 2.0502,
      "step": 20700
    },
    {
      "epoch": 0.4335681827656647,
      "grad_norm": 1.493910551071167,
      "learning_rate": 1.0839261054222363e-05,
      "loss": 2.0197,
      "step": 20800
    },
    {
      "epoch": 0.43565264518280733,
      "grad_norm": 1.8109139204025269,
      "learning_rate": 1.0891372886213815e-05,
      "loss": 2.055,
      "step": 20900
    },
    {
      "epoch": 0.43773710759995,
      "grad_norm": 1.8932719230651855,
      "learning_rate": 1.0943484718205268e-05,
      "loss": 2.0294,
      "step": 21000
    },
    {
      "epoch": 0.4398215700170926,
      "grad_norm": 1.4016826152801514,
      "learning_rate": 1.0995596550196723e-05,
      "loss": 2.0341,
      "step": 21100
    },
    {
      "epoch": 0.4419060324342352,
      "grad_norm": 1.1845966577529907,
      "learning_rate": 1.1047708382188175e-05,
      "loss": 2.04,
      "step": 21200
    },
    {
      "epoch": 0.4439904948513778,
      "grad_norm": 9.983181953430176,
      "learning_rate": 1.1099820214179631e-05,
      "loss": 2.0413,
      "step": 21300
    },
    {
      "epoch": 0.4460749572685204,
      "grad_norm": 1.4713338613510132,
      "learning_rate": 1.1151932046171083e-05,
      "loss": 2.0447,
      "step": 21400
    },
    {
      "epoch": 0.4481594196856631,
      "grad_norm": 1.1357747316360474,
      "learning_rate": 1.1204043878162539e-05,
      "loss": 2.008,
      "step": 21500
    },
    {
      "epoch": 0.4502438821028057,
      "grad_norm": 1.479264497756958,
      "learning_rate": 1.1256155710153991e-05,
      "loss": 2.0303,
      "step": 21600
    },
    {
      "epoch": 0.4523283445199483,
      "grad_norm": 1.3185663223266602,
      "learning_rate": 1.1308267542145447e-05,
      "loss": 2.0427,
      "step": 21700
    },
    {
      "epoch": 0.4544128069370909,
      "grad_norm": 1.1067243814468384,
      "learning_rate": 1.1360379374136899e-05,
      "loss": 2.0116,
      "step": 21800
    },
    {
      "epoch": 0.45649726935423357,
      "grad_norm": 1.9280306100845337,
      "learning_rate": 1.1412491206128353e-05,
      "loss": 2.0235,
      "step": 21900
    },
    {
      "epoch": 0.4585817317713762,
      "grad_norm": 1.7216917276382446,
      "learning_rate": 1.1464603038119807e-05,
      "loss": 2.0208,
      "step": 22000
    },
    {
      "epoch": 0.4606661941885188,
      "grad_norm": 1.0976884365081787,
      "learning_rate": 1.151671487011126e-05,
      "loss": 2.0185,
      "step": 22100
    },
    {
      "epoch": 0.4627506566056614,
      "grad_norm": 1.3677102327346802,
      "learning_rate": 1.1568826702102713e-05,
      "loss": 2.0132,
      "step": 22200
    },
    {
      "epoch": 0.464835119022804,
      "grad_norm": 1.1687935590744019,
      "learning_rate": 1.1620938534094167e-05,
      "loss": 2.0224,
      "step": 22300
    },
    {
      "epoch": 0.46691958143994666,
      "grad_norm": 2.3308753967285156,
      "learning_rate": 1.167305036608562e-05,
      "loss": 2.0277,
      "step": 22400
    },
    {
      "epoch": 0.46900404385708927,
      "grad_norm": 1.6247777938842773,
      "learning_rate": 1.1725162198077074e-05,
      "loss": 2.0265,
      "step": 22500
    },
    {
      "epoch": 0.4710885062742319,
      "grad_norm": 1.235781192779541,
      "learning_rate": 1.1777274030068528e-05,
      "loss": 2.0124,
      "step": 22600
    },
    {
      "epoch": 0.4731729686913745,
      "grad_norm": 1.9984397888183594,
      "learning_rate": 1.182938586205998e-05,
      "loss": 1.9896,
      "step": 22700
    },
    {
      "epoch": 0.4752574311085171,
      "grad_norm": 1.3742737770080566,
      "learning_rate": 1.1881497694051436e-05,
      "loss": 2.0,
      "step": 22800
    },
    {
      "epoch": 0.47734189352565976,
      "grad_norm": 2.8917479515075684,
      "learning_rate": 1.1933609526042888e-05,
      "loss": 2.0025,
      "step": 22900
    },
    {
      "epoch": 0.47942635594280236,
      "grad_norm": 1.2879999876022339,
      "learning_rate": 1.1985721358034344e-05,
      "loss": 2.0312,
      "step": 23000
    },
    {
      "epoch": 0.48151081835994497,
      "grad_norm": 0.9497489929199219,
      "learning_rate": 1.2037833190025796e-05,
      "loss": 2.0129,
      "step": 23100
    },
    {
      "epoch": 0.4835952807770876,
      "grad_norm": 1.982295274734497,
      "learning_rate": 1.2089945022017252e-05,
      "loss": 2.0036,
      "step": 23200
    },
    {
      "epoch": 0.4856797431942302,
      "grad_norm": 1.4037415981292725,
      "learning_rate": 1.2142056854008704e-05,
      "loss": 1.9977,
      "step": 23300
    },
    {
      "epoch": 0.48776420561137285,
      "grad_norm": 1.1049786806106567,
      "learning_rate": 1.2194168686000156e-05,
      "loss": 2.0144,
      "step": 23400
    },
    {
      "epoch": 0.48984866802851545,
      "grad_norm": 1.9810779094696045,
      "learning_rate": 1.2246280517991612e-05,
      "loss": 2.004,
      "step": 23500
    },
    {
      "epoch": 0.49193313044565806,
      "grad_norm": 1.4957938194274902,
      "learning_rate": 1.2298392349983064e-05,
      "loss": 1.9874,
      "step": 23600
    },
    {
      "epoch": 0.49401759286280067,
      "grad_norm": 1.414520263671875,
      "learning_rate": 1.235050418197452e-05,
      "loss": 2.0238,
      "step": 23700
    },
    {
      "epoch": 0.4961020552799433,
      "grad_norm": 1.2930928468704224,
      "learning_rate": 1.2402616013965972e-05,
      "loss": 1.9898,
      "step": 23800
    },
    {
      "epoch": 0.49818651769708594,
      "grad_norm": 1.3122200965881348,
      "learning_rate": 1.2454727845957426e-05,
      "loss": 1.9952,
      "step": 23900
    },
    {
      "epoch": 0.5002709801142285,
      "grad_norm": 3.088305950164795,
      "learning_rate": 1.250683967794888e-05,
      "loss": 1.9919,
      "step": 24000
    },
    {
      "epoch": 0.5023554425313712,
      "grad_norm": 1.2572028636932373,
      "learning_rate": 1.2558951509940333e-05,
      "loss": 2.0099,
      "step": 24100
    },
    {
      "epoch": 0.5044399049485138,
      "grad_norm": 1.3861281871795654,
      "learning_rate": 1.2611063341931786e-05,
      "loss": 2.0091,
      "step": 24200
    },
    {
      "epoch": 0.5065243673656564,
      "grad_norm": 3.073563814163208,
      "learning_rate": 1.2663175173923241e-05,
      "loss": 1.995,
      "step": 24300
    },
    {
      "epoch": 0.508608829782799,
      "grad_norm": 0.9809319972991943,
      "learning_rate": 1.2715287005914693e-05,
      "loss": 2.013,
      "step": 24400
    },
    {
      "epoch": 0.5106932921999416,
      "grad_norm": 1.243375301361084,
      "learning_rate": 1.2767398837906149e-05,
      "loss": 1.9997,
      "step": 24500
    },
    {
      "epoch": 0.5127777546170843,
      "grad_norm": 1.1517629623413086,
      "learning_rate": 1.2819510669897601e-05,
      "loss": 1.988,
      "step": 24600
    },
    {
      "epoch": 0.5148622170342269,
      "grad_norm": 1.890515685081482,
      "learning_rate": 1.2871622501889053e-05,
      "loss": 1.9872,
      "step": 24700
    },
    {
      "epoch": 0.5169466794513695,
      "grad_norm": 1.2368491888046265,
      "learning_rate": 1.2923734333880509e-05,
      "loss": 1.9788,
      "step": 24800
    },
    {
      "epoch": 0.5190311418685121,
      "grad_norm": 1.298607587814331,
      "learning_rate": 1.2975846165871961e-05,
      "loss": 1.9849,
      "step": 24900
    },
    {
      "epoch": 0.5211156042856547,
      "grad_norm": 1.2520874738693237,
      "learning_rate": 1.3027957997863417e-05,
      "loss": 2.0153,
      "step": 25000
    },
    {
      "epoch": 0.5232000667027974,
      "grad_norm": 1.6040500402450562,
      "learning_rate": 1.3080069829854869e-05,
      "loss": 1.9786,
      "step": 25100
    },
    {
      "epoch": 0.52528452911994,
      "grad_norm": 1.2517794370651245,
      "learning_rate": 1.3132181661846325e-05,
      "loss": 1.9779,
      "step": 25200
    },
    {
      "epoch": 0.5273689915370826,
      "grad_norm": 1.0364949703216553,
      "learning_rate": 1.3184293493837777e-05,
      "loss": 2.001,
      "step": 25300
    },
    {
      "epoch": 0.5294534539542252,
      "grad_norm": 1.2333202362060547,
      "learning_rate": 1.323640532582923e-05,
      "loss": 2.0127,
      "step": 25400
    },
    {
      "epoch": 0.5315379163713678,
      "grad_norm": 1.1749173402786255,
      "learning_rate": 1.3288517157820685e-05,
      "loss": 1.9699,
      "step": 25500
    },
    {
      "epoch": 0.5336223787885105,
      "grad_norm": 1.2822164297103882,
      "learning_rate": 1.3340628989812138e-05,
      "loss": 1.9742,
      "step": 25600
    },
    {
      "epoch": 0.535706841205653,
      "grad_norm": 12.977947235107422,
      "learning_rate": 1.3392740821803592e-05,
      "loss": 1.9848,
      "step": 25700
    },
    {
      "epoch": 0.5377913036227957,
      "grad_norm": 1.4702880382537842,
      "learning_rate": 1.3444852653795046e-05,
      "loss": 1.9828,
      "step": 25800
    },
    {
      "epoch": 0.5398757660399383,
      "grad_norm": 1.9451568126678467,
      "learning_rate": 1.3496964485786498e-05,
      "loss": 1.9818,
      "step": 25900
    },
    {
      "epoch": 0.5419602284570809,
      "grad_norm": 1.347174882888794,
      "learning_rate": 1.3549076317777952e-05,
      "loss": 1.9821,
      "step": 26000
    },
    {
      "epoch": 0.5440446908742236,
      "grad_norm": 1.4248614311218262,
      "learning_rate": 1.3601188149769406e-05,
      "loss": 1.9753,
      "step": 26100
    },
    {
      "epoch": 0.5461291532913661,
      "grad_norm": 1.8619277477264404,
      "learning_rate": 1.3653299981760858e-05,
      "loss": 1.9882,
      "step": 26200
    },
    {
      "epoch": 0.5482136157085088,
      "grad_norm": 1.6204618215560913,
      "learning_rate": 1.3705411813752314e-05,
      "loss": 1.9807,
      "step": 26300
    },
    {
      "epoch": 0.5502980781256513,
      "grad_norm": 1.8339996337890625,
      "learning_rate": 1.3757523645743766e-05,
      "loss": 1.9563,
      "step": 26400
    },
    {
      "epoch": 0.552382540542794,
      "grad_norm": 1.0893349647521973,
      "learning_rate": 1.3809635477735222e-05,
      "loss": 1.9731,
      "step": 26500
    },
    {
      "epoch": 0.5544670029599367,
      "grad_norm": 1.1871613264083862,
      "learning_rate": 1.3861747309726674e-05,
      "loss": 1.9645,
      "step": 26600
    },
    {
      "epoch": 0.5565514653770792,
      "grad_norm": 1.064344882965088,
      "learning_rate": 1.391385914171813e-05,
      "loss": 1.9775,
      "step": 26700
    },
    {
      "epoch": 0.5586359277942219,
      "grad_norm": 0.9557547569274902,
      "learning_rate": 1.3965970973709582e-05,
      "loss": 1.9658,
      "step": 26800
    },
    {
      "epoch": 0.5607203902113644,
      "grad_norm": 1.0063979625701904,
      "learning_rate": 1.4018082805701037e-05,
      "loss": 1.9578,
      "step": 26900
    },
    {
      "epoch": 0.5628048526285071,
      "grad_norm": 1.4233479499816895,
      "learning_rate": 1.407019463769249e-05,
      "loss": 1.971,
      "step": 27000
    },
    {
      "epoch": 0.5648893150456498,
      "grad_norm": 1.1650019884109497,
      "learning_rate": 1.4122306469683942e-05,
      "loss": 1.9632,
      "step": 27100
    },
    {
      "epoch": 0.5669737774627923,
      "grad_norm": 1.025070309638977,
      "learning_rate": 1.4174418301675397e-05,
      "loss": 1.9593,
      "step": 27200
    },
    {
      "epoch": 0.569058239879935,
      "grad_norm": 0.8504909873008728,
      "learning_rate": 1.422653013366685e-05,
      "loss": 1.9738,
      "step": 27300
    },
    {
      "epoch": 0.5711427022970775,
      "grad_norm": 1.2217037677764893,
      "learning_rate": 1.4278641965658303e-05,
      "loss": 1.9674,
      "step": 27400
    },
    {
      "epoch": 0.5732271647142202,
      "grad_norm": 1.1072654724121094,
      "learning_rate": 1.4330753797649757e-05,
      "loss": 1.9591,
      "step": 27500
    },
    {
      "epoch": 0.5753116271313629,
      "grad_norm": 1.6574842929840088,
      "learning_rate": 1.4382865629641211e-05,
      "loss": 1.9743,
      "step": 27600
    },
    {
      "epoch": 0.5773960895485054,
      "grad_norm": 1.6935937404632568,
      "learning_rate": 1.4434977461632665e-05,
      "loss": 1.9584,
      "step": 27700
    },
    {
      "epoch": 0.5794805519656481,
      "grad_norm": 0.9915087819099426,
      "learning_rate": 1.4487089293624119e-05,
      "loss": 1.9661,
      "step": 27800
    },
    {
      "epoch": 0.5815650143827906,
      "grad_norm": 1.6524442434310913,
      "learning_rate": 1.4539201125615571e-05,
      "loss": 1.9888,
      "step": 27900
    },
    {
      "epoch": 0.5836494767999333,
      "grad_norm": 0.9779165983200073,
      "learning_rate": 1.4591312957607027e-05,
      "loss": 1.9528,
      "step": 28000
    },
    {
      "epoch": 0.585733939217076,
      "grad_norm": 1.5924351215362549,
      "learning_rate": 1.4643424789598479e-05,
      "loss": 1.9434,
      "step": 28100
    },
    {
      "epoch": 0.5878184016342185,
      "grad_norm": 1.073769211769104,
      "learning_rate": 1.4695536621589935e-05,
      "loss": 1.9859,
      "step": 28200
    },
    {
      "epoch": 0.5899028640513612,
      "grad_norm": 1.2847269773483276,
      "learning_rate": 1.4747648453581387e-05,
      "loss": 1.9691,
      "step": 28300
    },
    {
      "epoch": 0.5919873264685038,
      "grad_norm": 2.1441261768341064,
      "learning_rate": 1.4799760285572839e-05,
      "loss": 1.9548,
      "step": 28400
    },
    {
      "epoch": 0.5940717888856464,
      "grad_norm": 1.0148779153823853,
      "learning_rate": 1.4851872117564295e-05,
      "loss": 1.956,
      "step": 28500
    },
    {
      "epoch": 0.596156251302789,
      "grad_norm": 1.3669894933700562,
      "learning_rate": 1.4903983949555747e-05,
      "loss": 1.9456,
      "step": 28600
    },
    {
      "epoch": 0.5982407137199316,
      "grad_norm": 2.084087610244751,
      "learning_rate": 1.4956095781547202e-05,
      "loss": 1.9768,
      "step": 28700
    },
    {
      "epoch": 0.6003251761370743,
      "grad_norm": 1.0451395511627197,
      "learning_rate": 1.5008207613538655e-05,
      "loss": 1.9628,
      "step": 28800
    },
    {
      "epoch": 0.6024096385542169,
      "grad_norm": 1.0861454010009766,
      "learning_rate": 1.5060319445530109e-05,
      "loss": 1.9608,
      "step": 28900
    },
    {
      "epoch": 0.6044941009713595,
      "grad_norm": 1.082926630973816,
      "learning_rate": 1.5112431277521562e-05,
      "loss": 1.9507,
      "step": 29000
    },
    {
      "epoch": 0.6065785633885021,
      "grad_norm": 1.0375463962554932,
      "learning_rate": 1.5164543109513016e-05,
      "loss": 1.9444,
      "step": 29100
    },
    {
      "epoch": 0.6086630258056447,
      "grad_norm": 1.0970326662063599,
      "learning_rate": 1.521665494150447e-05,
      "loss": 1.9602,
      "step": 29200
    },
    {
      "epoch": 0.6107474882227873,
      "grad_norm": 1.4305795431137085,
      "learning_rate": 1.5268766773495924e-05,
      "loss": 1.9663,
      "step": 29300
    },
    {
      "epoch": 0.61283195063993,
      "grad_norm": 1.2782924175262451,
      "learning_rate": 1.5320878605487378e-05,
      "loss": 1.9706,
      "step": 29400
    },
    {
      "epoch": 0.6149164130570726,
      "grad_norm": 1.2696731090545654,
      "learning_rate": 1.5372990437478832e-05,
      "loss": 1.9502,
      "step": 29500
    },
    {
      "epoch": 0.6170008754742152,
      "grad_norm": 1.0566843748092651,
      "learning_rate": 1.5425102269470286e-05,
      "loss": 1.948,
      "step": 29600
    },
    {
      "epoch": 0.6190853378913578,
      "grad_norm": 1.1071428060531616,
      "learning_rate": 1.5477214101461736e-05,
      "loss": 1.9409,
      "step": 29700
    },
    {
      "epoch": 0.6211698003085004,
      "grad_norm": 0.9255651235580444,
      "learning_rate": 1.5529325933453194e-05,
      "loss": 1.9466,
      "step": 29800
    },
    {
      "epoch": 0.6232542627256431,
      "grad_norm": 1.0186619758605957,
      "learning_rate": 1.5581437765444644e-05,
      "loss": 1.9493,
      "step": 29900
    },
    {
      "epoch": 0.6253387251427857,
      "grad_norm": 1.451917052268982,
      "learning_rate": 1.5633549597436098e-05,
      "loss": 1.9503,
      "step": 30000
    },
    {
      "epoch": 0.6274231875599283,
      "grad_norm": 1.0187926292419434,
      "learning_rate": 1.5685661429427552e-05,
      "loss": 1.9563,
      "step": 30100
    },
    {
      "epoch": 0.6295076499770709,
      "grad_norm": 0.871369481086731,
      "learning_rate": 1.5737773261419006e-05,
      "loss": 1.9549,
      "step": 30200
    },
    {
      "epoch": 0.6315921123942135,
      "grad_norm": 0.98649662733078,
      "learning_rate": 1.578988509341046e-05,
      "loss": 1.9596,
      "step": 30300
    },
    {
      "epoch": 0.6336765748113562,
      "grad_norm": 1.6009703874588013,
      "learning_rate": 1.5841996925401914e-05,
      "loss": 1.9348,
      "step": 30400
    },
    {
      "epoch": 0.6357610372284987,
      "grad_norm": 1.0804214477539062,
      "learning_rate": 1.5894108757393368e-05,
      "loss": 1.9428,
      "step": 30500
    },
    {
      "epoch": 0.6378454996456414,
      "grad_norm": 1.1061315536499023,
      "learning_rate": 1.594622058938482e-05,
      "loss": 1.9371,
      "step": 30600
    },
    {
      "epoch": 0.639929962062784,
      "grad_norm": 1.0033613443374634,
      "learning_rate": 1.5998332421376275e-05,
      "loss": 1.9508,
      "step": 30700
    },
    {
      "epoch": 0.6420144244799266,
      "grad_norm": 0.9282180070877075,
      "learning_rate": 1.6050444253367726e-05,
      "loss": 1.9376,
      "step": 30800
    },
    {
      "epoch": 0.6440988868970693,
      "grad_norm": 0.9301239252090454,
      "learning_rate": 1.6102556085359183e-05,
      "loss": 1.9556,
      "step": 30900
    },
    {
      "epoch": 0.6461833493142118,
      "grad_norm": 0.9229454398155212,
      "learning_rate": 1.6154667917350634e-05,
      "loss": 1.9314,
      "step": 31000
    },
    {
      "epoch": 0.6482678117313545,
      "grad_norm": 1.0757591724395752,
      "learning_rate": 1.620677974934209e-05,
      "loss": 1.9309,
      "step": 31100
    },
    {
      "epoch": 0.6503522741484971,
      "grad_norm": 3.362175464630127,
      "learning_rate": 1.625889158133354e-05,
      "loss": 1.9254,
      "step": 31200
    },
    {
      "epoch": 0.6524367365656397,
      "grad_norm": 1.1901942491531372,
      "learning_rate": 1.6311003413325e-05,
      "loss": 1.9337,
      "step": 31300
    },
    {
      "epoch": 0.6545211989827824,
      "grad_norm": 1.2026360034942627,
      "learning_rate": 1.636311524531645e-05,
      "loss": 1.9407,
      "step": 31400
    },
    {
      "epoch": 0.6566056613999249,
      "grad_norm": 1.247865915298462,
      "learning_rate": 1.6415227077307903e-05,
      "loss": 1.9345,
      "step": 31500
    },
    {
      "epoch": 0.6586901238170676,
      "grad_norm": 1.0052815675735474,
      "learning_rate": 1.6467338909299357e-05,
      "loss": 1.9295,
      "step": 31600
    },
    {
      "epoch": 0.6607745862342101,
      "grad_norm": 0.9365499019622803,
      "learning_rate": 1.651945074129081e-05,
      "loss": 1.9421,
      "step": 31700
    },
    {
      "epoch": 0.6628590486513528,
      "grad_norm": 1.7018957138061523,
      "learning_rate": 1.6571562573282265e-05,
      "loss": 1.9174,
      "step": 31800
    },
    {
      "epoch": 0.6649435110684955,
      "grad_norm": 1.0856178998947144,
      "learning_rate": 1.662367440527372e-05,
      "loss": 1.9045,
      "step": 31900
    },
    {
      "epoch": 0.667027973485638,
      "grad_norm": 1.111540675163269,
      "learning_rate": 1.6675786237265173e-05,
      "loss": 1.9263,
      "step": 32000
    },
    {
      "epoch": 0.6691124359027807,
      "grad_norm": 0.9781438708305359,
      "learning_rate": 1.6727898069256626e-05,
      "loss": 1.9115,
      "step": 32100
    },
    {
      "epoch": 0.6711968983199232,
      "grad_norm": 0.9509501457214355,
      "learning_rate": 1.678000990124808e-05,
      "loss": 1.9194,
      "step": 32200
    },
    {
      "epoch": 0.6732813607370659,
      "grad_norm": 1.0077457427978516,
      "learning_rate": 1.6832121733239534e-05,
      "loss": 1.9361,
      "step": 32300
    },
    {
      "epoch": 0.6753658231542086,
      "grad_norm": 1.0101159811019897,
      "learning_rate": 1.6884233565230988e-05,
      "loss": 1.9227,
      "step": 32400
    },
    {
      "epoch": 0.6774502855713511,
      "grad_norm": 1.6001538038253784,
      "learning_rate": 1.693634539722244e-05,
      "loss": 1.9114,
      "step": 32500
    },
    {
      "epoch": 0.6795347479884938,
      "grad_norm": 1.300024151802063,
      "learning_rate": 1.6988457229213896e-05,
      "loss": 1.9471,
      "step": 32600
    },
    {
      "epoch": 0.6816192104056363,
      "grad_norm": 1.0062751770019531,
      "learning_rate": 1.7040569061205346e-05,
      "loss": 1.9043,
      "step": 32700
    },
    {
      "epoch": 0.683703672822779,
      "grad_norm": 1.012172818183899,
      "learning_rate": 1.7092680893196804e-05,
      "loss": 1.9189,
      "step": 32800
    },
    {
      "epoch": 0.6857881352399217,
      "grad_norm": 2.5829713344573975,
      "learning_rate": 1.7144792725188254e-05,
      "loss": 1.8859,
      "step": 32900
    },
    {
      "epoch": 0.6878725976570642,
      "grad_norm": 1.1416168212890625,
      "learning_rate": 1.719690455717971e-05,
      "loss": 1.9426,
      "step": 33000
    },
    {
      "epoch": 0.6899570600742069,
      "grad_norm": 5.545751094818115,
      "learning_rate": 1.7249016389171162e-05,
      "loss": 1.9544,
      "step": 33100
    },
    {
      "epoch": 0.6920415224913494,
      "grad_norm": 1.051398515701294,
      "learning_rate": 1.7301128221162616e-05,
      "loss": 1.9391,
      "step": 33200
    },
    {
      "epoch": 0.6941259849084921,
      "grad_norm": 1.153196930885315,
      "learning_rate": 1.735324005315407e-05,
      "loss": 1.9112,
      "step": 33300
    },
    {
      "epoch": 0.6962104473256348,
      "grad_norm": 1.1652779579162598,
      "learning_rate": 1.7405351885145524e-05,
      "loss": 1.9035,
      "step": 33400
    },
    {
      "epoch": 0.6982949097427773,
      "grad_norm": 1.9276180267333984,
      "learning_rate": 1.7457463717136978e-05,
      "loss": 1.9415,
      "step": 33500
    },
    {
      "epoch": 0.70037937215992,
      "grad_norm": 0.8301405310630798,
      "learning_rate": 1.750957554912843e-05,
      "loss": 1.9054,
      "step": 33600
    },
    {
      "epoch": 0.7024638345770626,
      "grad_norm": 0.9777198433876038,
      "learning_rate": 1.7561687381119885e-05,
      "loss": 1.8924,
      "step": 33700
    },
    {
      "epoch": 0.7045482969942052,
      "grad_norm": 1.2608678340911865,
      "learning_rate": 1.761379921311134e-05,
      "loss": 1.8982,
      "step": 33800
    },
    {
      "epoch": 0.7066327594113478,
      "grad_norm": 0.9381488561630249,
      "learning_rate": 1.7665911045102793e-05,
      "loss": 1.89,
      "step": 33900
    },
    {
      "epoch": 0.7087172218284904,
      "grad_norm": 1.045607328414917,
      "learning_rate": 1.7718022877094244e-05,
      "loss": 1.9073,
      "step": 34000
    },
    {
      "epoch": 0.7108016842456331,
      "grad_norm": 1.0266004800796509,
      "learning_rate": 1.77701347090857e-05,
      "loss": 1.9222,
      "step": 34100
    },
    {
      "epoch": 0.7128861466627757,
      "grad_norm": 1.1733341217041016,
      "learning_rate": 1.782224654107715e-05,
      "loss": 1.9041,
      "step": 34200
    },
    {
      "epoch": 0.7149706090799183,
      "grad_norm": 1.0179702043533325,
      "learning_rate": 1.787435837306861e-05,
      "loss": 1.9133,
      "step": 34300
    },
    {
      "epoch": 0.7170550714970609,
      "grad_norm": 1.2114508152008057,
      "learning_rate": 1.792647020506006e-05,
      "loss": 1.8858,
      "step": 34400
    },
    {
      "epoch": 0.7191395339142035,
      "grad_norm": 0.9730684161186218,
      "learning_rate": 1.7978582037051513e-05,
      "loss": 1.9143,
      "step": 34500
    },
    {
      "epoch": 0.7212239963313462,
      "grad_norm": 0.9815762639045715,
      "learning_rate": 1.8030693869042967e-05,
      "loss": 1.9032,
      "step": 34600
    },
    {
      "epoch": 0.7233084587484888,
      "grad_norm": 0.979191243648529,
      "learning_rate": 1.808280570103442e-05,
      "loss": 1.8858,
      "step": 34700
    },
    {
      "epoch": 0.7253929211656314,
      "grad_norm": 1.051487684249878,
      "learning_rate": 1.8134917533025875e-05,
      "loss": 1.8935,
      "step": 34800
    },
    {
      "epoch": 0.727477383582774,
      "grad_norm": 1.1583081483840942,
      "learning_rate": 1.818702936501733e-05,
      "loss": 1.8998,
      "step": 34900
    },
    {
      "epoch": 0.7295618459999166,
      "grad_norm": 1.3905084133148193,
      "learning_rate": 1.8239141197008783e-05,
      "loss": 1.9105,
      "step": 35000
    },
    {
      "epoch": 0.7316463084170592,
      "grad_norm": 0.9211100339889526,
      "learning_rate": 1.8291253029000237e-05,
      "loss": 1.9145,
      "step": 35100
    },
    {
      "epoch": 0.7337307708342019,
      "grad_norm": 0.9945178627967834,
      "learning_rate": 1.834336486099169e-05,
      "loss": 1.9049,
      "step": 35200
    },
    {
      "epoch": 0.7358152332513445,
      "grad_norm": 0.9826840162277222,
      "learning_rate": 1.8395476692983144e-05,
      "loss": 1.8968,
      "step": 35300
    },
    {
      "epoch": 0.7378996956684871,
      "grad_norm": 1.0781490802764893,
      "learning_rate": 1.8447588524974598e-05,
      "loss": 1.9173,
      "step": 35400
    },
    {
      "epoch": 0.7399841580856297,
      "grad_norm": 1.347640037536621,
      "learning_rate": 1.849970035696605e-05,
      "loss": 1.8976,
      "step": 35500
    },
    {
      "epoch": 0.7420686205027723,
      "grad_norm": 0.9135119318962097,
      "learning_rate": 1.8551812188957506e-05,
      "loss": 1.8965,
      "step": 35600
    },
    {
      "epoch": 0.744153082919915,
      "grad_norm": 1.1346653699874878,
      "learning_rate": 1.8603924020948957e-05,
      "loss": 1.8989,
      "step": 35700
    },
    {
      "epoch": 0.7462375453370576,
      "grad_norm": 1.0951207876205444,
      "learning_rate": 1.865603585294041e-05,
      "loss": 1.8982,
      "step": 35800
    },
    {
      "epoch": 0.7483220077542002,
      "grad_norm": 1.0009814500808716,
      "learning_rate": 1.8708147684931864e-05,
      "loss": 1.8914,
      "step": 35900
    },
    {
      "epoch": 0.7504064701713428,
      "grad_norm": 0.9622611403465271,
      "learning_rate": 1.8760259516923318e-05,
      "loss": 1.8844,
      "step": 36000
    },
    {
      "epoch": 0.7524909325884854,
      "grad_norm": 1.4782415628433228,
      "learning_rate": 1.8812371348914772e-05,
      "loss": 1.8879,
      "step": 36100
    },
    {
      "epoch": 0.7545753950056281,
      "grad_norm": 1.0756487846374512,
      "learning_rate": 1.8864483180906226e-05,
      "loss": 1.908,
      "step": 36200
    },
    {
      "epoch": 0.7566598574227706,
      "grad_norm": 1.1616325378417969,
      "learning_rate": 1.891659501289768e-05,
      "loss": 1.9068,
      "step": 36300
    },
    {
      "epoch": 0.7587443198399133,
      "grad_norm": 1.1253652572631836,
      "learning_rate": 1.8968706844889134e-05,
      "loss": 1.9044,
      "step": 36400
    },
    {
      "epoch": 0.7608287822570559,
      "grad_norm": 0.9540199041366577,
      "learning_rate": 1.9020818676880588e-05,
      "loss": 1.857,
      "step": 36500
    },
    {
      "epoch": 0.7629132446741985,
      "grad_norm": 1.3189705610275269,
      "learning_rate": 1.907293050887204e-05,
      "loss": 1.8842,
      "step": 36600
    },
    {
      "epoch": 0.7649977070913412,
      "grad_norm": 1.7837036848068237,
      "learning_rate": 1.9125042340863496e-05,
      "loss": 1.861,
      "step": 36700
    },
    {
      "epoch": 0.7670821695084837,
      "grad_norm": 1.01956307888031,
      "learning_rate": 1.917715417285495e-05,
      "loss": 1.8852,
      "step": 36800
    },
    {
      "epoch": 0.7691666319256264,
      "grad_norm": 1.1640490293502808,
      "learning_rate": 1.9229266004846403e-05,
      "loss": 1.8695,
      "step": 36900
    },
    {
      "epoch": 0.771251094342769,
      "grad_norm": 1.3748284578323364,
      "learning_rate": 1.9281377836837857e-05,
      "loss": 1.893,
      "step": 37000
    },
    {
      "epoch": 0.7733355567599116,
      "grad_norm": 0.8749508261680603,
      "learning_rate": 1.9333489668829308e-05,
      "loss": 1.8791,
      "step": 37100
    },
    {
      "epoch": 0.7754200191770543,
      "grad_norm": 0.9299560785293579,
      "learning_rate": 1.938560150082076e-05,
      "loss": 1.8808,
      "step": 37200
    },
    {
      "epoch": 0.7775044815941968,
      "grad_norm": 0.8478776216506958,
      "learning_rate": 1.9437713332812216e-05,
      "loss": 1.8827,
      "step": 37300
    },
    {
      "epoch": 0.7795889440113395,
      "grad_norm": 1.0107454061508179,
      "learning_rate": 1.948982516480367e-05,
      "loss": 1.9092,
      "step": 37400
    },
    {
      "epoch": 0.781673406428482,
      "grad_norm": 1.0087549686431885,
      "learning_rate": 1.9541936996795123e-05,
      "loss": 1.8792,
      "step": 37500
    },
    {
      "epoch": 0.7837578688456247,
      "grad_norm": 1.1416882276535034,
      "learning_rate": 1.9594048828786577e-05,
      "loss": 1.8796,
      "step": 37600
    },
    {
      "epoch": 0.7858423312627674,
      "grad_norm": 0.9162054657936096,
      "learning_rate": 1.964616066077803e-05,
      "loss": 1.8829,
      "step": 37700
    },
    {
      "epoch": 0.7879267936799099,
      "grad_norm": 0.990573525428772,
      "learning_rate": 1.9698272492769485e-05,
      "loss": 1.8859,
      "step": 37800
    },
    {
      "epoch": 0.7900112560970526,
      "grad_norm": 0.8790944814682007,
      "learning_rate": 1.975038432476094e-05,
      "loss": 1.8789,
      "step": 37900
    },
    {
      "epoch": 0.7920957185141951,
      "grad_norm": 2.4909720420837402,
      "learning_rate": 1.9802496156752393e-05,
      "loss": 1.9049,
      "step": 38000
    },
    {
      "epoch": 0.7941801809313378,
      "grad_norm": 1.4410492181777954,
      "learning_rate": 1.9854607988743847e-05,
      "loss": 1.8903,
      "step": 38100
    },
    {
      "epoch": 0.7962646433484805,
      "grad_norm": 1.854480266571045,
      "learning_rate": 1.9906719820735297e-05,
      "loss": 1.8948,
      "step": 38200
    },
    {
      "epoch": 0.798349105765623,
      "grad_norm": 0.9377720952033997,
      "learning_rate": 1.9958831652726755e-05,
      "loss": 1.8835,
      "step": 38300
    },
    {
      "epoch": 0.8004335681827657,
      "grad_norm": 0.863426148891449,
      "learning_rate": 1.9999763616765983e-05,
      "loss": 1.888,
      "step": 38400
    },
    {
      "epoch": 0.8025180305999083,
      "grad_norm": 0.884555459022522,
      "learning_rate": 1.9992153179340215e-05,
      "loss": 1.8643,
      "step": 38500
    },
    {
      "epoch": 0.8046024930170509,
      "grad_norm": 1.795231819152832,
      "learning_rate": 1.9973831740621872e-05,
      "loss": 1.8671,
      "step": 38600
    },
    {
      "epoch": 0.8066869554341936,
      "grad_norm": 0.8534547090530396,
      "learning_rate": 1.9944818940117093e-05,
      "loss": 1.8869,
      "step": 38700
    },
    {
      "epoch": 0.8087714178513361,
      "grad_norm": 0.8905624747276306,
      "learning_rate": 1.9905145877843883e-05,
      "loss": 1.8813,
      "step": 38800
    },
    {
      "epoch": 0.8108558802684788,
      "grad_norm": 1.0225939750671387,
      "learning_rate": 1.9854855080994695e-05,
      "loss": 1.8797,
      "step": 38900
    },
    {
      "epoch": 0.8129403426856214,
      "grad_norm": 1.0849330425262451,
      "learning_rate": 1.9794000458349786e-05,
      "loss": 1.8685,
      "step": 39000
    },
    {
      "epoch": 0.815024805102764,
      "grad_norm": 1.1683331727981567,
      "learning_rate": 1.9722647242490173e-05,
      "loss": 1.8658,
      "step": 39100
    },
    {
      "epoch": 0.8171092675199066,
      "grad_norm": 2.3962318897247314,
      "learning_rate": 1.9640871919872137e-05,
      "loss": 1.8733,
      "step": 39200
    },
    {
      "epoch": 0.8191937299370492,
      "grad_norm": 0.8535931706428528,
      "learning_rate": 1.9548762148838235e-05,
      "loss": 1.8777,
      "step": 39300
    },
    {
      "epoch": 0.8212781923541919,
      "grad_norm": 1.10763418674469,
      "learning_rate": 1.94464166656527e-05,
      "loss": 1.8821,
      "step": 39400
    },
    {
      "epoch": 0.8233626547713345,
      "grad_norm": 5.269350528717041,
      "learning_rate": 1.933394517866198e-05,
      "loss": 1.8522,
      "step": 39500
    },
    {
      "epoch": 0.8254471171884771,
      "grad_norm": 1.2906358242034912,
      "learning_rate": 1.9211468250693814e-05,
      "loss": 1.8851,
      "step": 39600
    },
    {
      "epoch": 0.8275315796056197,
      "grad_norm": 1.466866374015808,
      "learning_rate": 1.907911716982097e-05,
      "loss": 1.8516,
      "step": 39700
    },
    {
      "epoch": 0.8296160420227623,
      "grad_norm": 0.8515104651451111,
      "learning_rate": 1.8937033808628157e-05,
      "loss": 1.8722,
      "step": 39800
    },
    {
      "epoch": 0.831700504439905,
      "grad_norm": 0.8316900134086609,
      "learning_rate": 1.8785370472132925e-05,
      "loss": 1.8794,
      "step": 39900
    },
    {
      "epoch": 0.8337849668570476,
      "grad_norm": 2.936185598373413,
      "learning_rate": 1.8624289734523655e-05,
      "loss": 1.8766,
      "step": 40000
    },
    {
      "epoch": 0.8358694292741902,
      "grad_norm": 0.8948720693588257,
      "learning_rate": 1.8453964264889555e-05,
      "loss": 1.8679,
      "step": 40100
    },
    {
      "epoch": 0.8379538916913328,
      "grad_norm": 1.0251227617263794,
      "learning_rate": 1.8274576642129552e-05,
      "loss": 1.8599,
      "step": 40200
    },
    {
      "epoch": 0.8400383541084754,
      "grad_norm": 2.930690050125122,
      "learning_rate": 1.8086319159238442e-05,
      "loss": 1.8634,
      "step": 40300
    },
    {
      "epoch": 0.842122816525618,
      "grad_norm": 1.140226125717163,
      "learning_rate": 1.7889393617180074e-05,
      "loss": 1.8637,
      "step": 40400
    },
    {
      "epoch": 0.8442072789427607,
      "grad_norm": 0.9171246290206909,
      "learning_rate": 1.7684011108568593e-05,
      "loss": 1.854,
      "step": 40500
    },
    {
      "epoch": 0.8462917413599033,
      "grad_norm": 0.9319982528686523,
      "learning_rate": 1.747039179138954e-05,
      "loss": 1.8384,
      "step": 40600
    },
    {
      "epoch": 0.8483762037770459,
      "grad_norm": 1.0997920036315918,
      "learning_rate": 1.7248764653003433e-05,
      "loss": 1.8613,
      "step": 40700
    },
    {
      "epoch": 0.8504606661941885,
      "grad_norm": 0.8980279564857483,
      "learning_rate": 1.701936726468477e-05,
      "loss": 1.8864,
      "step": 40800
    },
    {
      "epoch": 0.8525451286113311,
      "grad_norm": 1.1140270233154297,
      "learning_rate": 1.6782445526959553e-05,
      "loss": 1.8564,
      "step": 40900
    },
    {
      "epoch": 0.8546295910284738,
      "grad_norm": 0.9438982009887695,
      "learning_rate": 1.65382534060144e-05,
      "loss": 1.8686,
      "step": 41000
    },
    {
      "epoch": 0.8567140534456164,
      "grad_norm": 1.087593913078308,
      "learning_rate": 1.628705266145969e-05,
      "loss": 1.8612,
      "step": 41100
    },
    {
      "epoch": 0.858798515862759,
      "grad_norm": 1.5680657625198364,
      "learning_rate": 1.6029112565738647e-05,
      "loss": 1.8595,
      "step": 41200
    },
    {
      "epoch": 0.8608829782799016,
      "grad_norm": 1.1920835971832275,
      "learning_rate": 1.5764709615483084e-05,
      "loss": 1.8368,
      "step": 41300
    },
    {
      "epoch": 0.8629674406970442,
      "grad_norm": 0.9491363167762756,
      "learning_rate": 1.549412723512526e-05,
      "loss": 1.8433,
      "step": 41400
    },
    {
      "epoch": 0.8650519031141869,
      "grad_norm": 0.8679408431053162,
      "learning_rate": 1.521765547308355e-05,
      "loss": 1.8585,
      "step": 41500
    },
    {
      "epoch": 0.8671363655313294,
      "grad_norm": 0.9135258793830872,
      "learning_rate": 1.4935590690847555e-05,
      "loss": 1.8535,
      "step": 41600
    },
    {
      "epoch": 0.8692208279484721,
      "grad_norm": 0.9444493055343628,
      "learning_rate": 1.464823524529601e-05,
      "loss": 1.8606,
      "step": 41700
    },
    {
      "epoch": 0.8713052903656147,
      "grad_norm": 0.9098984003067017,
      "learning_rate": 1.4355897164587957e-05,
      "loss": 1.8634,
      "step": 41800
    },
    {
      "epoch": 0.8733897527827573,
      "grad_norm": 4.627810478210449,
      "learning_rate": 1.4058889817974664e-05,
      "loss": 1.8349,
      "step": 41900
    },
    {
      "epoch": 0.8754742151999,
      "grad_norm": 0.8299252986907959,
      "learning_rate": 1.3757531579886186e-05,
      "loss": 1.8541,
      "step": 42000
    },
    {
      "epoch": 0.8775586776170425,
      "grad_norm": 0.9095094203948975,
      "learning_rate": 1.345214548865267e-05,
      "loss": 1.885,
      "step": 42100
    },
    {
      "epoch": 0.8796431400341852,
      "grad_norm": 0.9644302725791931,
      "learning_rate": 1.3143058900226262e-05,
      "loss": 1.8788,
      "step": 42200
    },
    {
      "epoch": 0.8817276024513278,
      "grad_norm": 1.4231350421905518,
      "learning_rate": 1.2830603137274729e-05,
      "loss": 1.8451,
      "step": 42300
    },
    {
      "epoch": 0.8838120648684704,
      "grad_norm": 1.0535882711410522,
      "learning_rate": 1.2515113134023035e-05,
      "loss": 1.864,
      "step": 42400
    },
    {
      "epoch": 0.8858965272856131,
      "grad_norm": 1.3644357919692993,
      "learning_rate": 1.2196927077223523e-05,
      "loss": 1.853,
      "step": 42500
    },
    {
      "epoch": 0.8879809897027556,
      "grad_norm": 0.9575700163841248,
      "learning_rate": 1.187638604363958e-05,
      "loss": 1.851,
      "step": 42600
    },
    {
      "epoch": 0.8900654521198983,
      "grad_norm": 1.0932910442352295,
      "learning_rate": 1.1553833634431375e-05,
      "loss": 1.8539,
      "step": 42700
    },
    {
      "epoch": 0.8921499145370408,
      "grad_norm": 0.8120511174201965,
      "learning_rate": 1.1229615606835609e-05,
      "loss": 1.8331,
      "step": 42800
    },
    {
      "epoch": 0.8942343769541835,
      "grad_norm": 0.967151939868927,
      "learning_rate": 1.0904079503534057e-05,
      "loss": 1.8565,
      "step": 42900
    },
    {
      "epoch": 0.8963188393713262,
      "grad_norm": 0.8279860019683838,
      "learning_rate": 1.0577574280108226e-05,
      "loss": 1.8469,
      "step": 43000
    },
    {
      "epoch": 0.8984033017884687,
      "grad_norm": 0.986449658870697,
      "learning_rate": 1.0250449930979484e-05,
      "loss": 1.8506,
      "step": 43100
    },
    {
      "epoch": 0.9004877642056114,
      "grad_norm": 0.9188694953918457,
      "learning_rate": 9.92305711423557e-06,
      "loss": 1.8345,
      "step": 43200
    },
    {
      "epoch": 0.9025722266227539,
      "grad_norm": 0.8622938990592957,
      "learning_rate": 9.595746775745753e-06,
      "loss": 1.8451,
      "step": 43300
    },
    {
      "epoch": 0.9046566890398966,
      "grad_norm": 1.5604217052459717,
      "learning_rate": 9.268869772967448e-06,
      "loss": 1.8323,
      "step": 43400
    },
    {
      "epoch": 0.9067411514570393,
      "grad_norm": 0.8539250493049622,
      "learning_rate": 8.942776498847645e-06,
      "loss": 1.8514,
      "step": 43500
    },
    {
      "epoch": 0.9088256138741818,
      "grad_norm": 0.989700973033905,
      "learning_rate": 8.617816506222246e-06,
      "loss": 1.8386,
      "step": 43600
    },
    {
      "epoch": 0.9109100762913245,
      "grad_norm": 0.7901672124862671,
      "learning_rate": 8.294338133115956e-06,
      "loss": 1.8395,
      "step": 43700
    },
    {
      "epoch": 0.9129945387084671,
      "grad_norm": 0.9410315155982971,
      "learning_rate": 7.972688129344382e-06,
      "loss": 1.8545,
      "step": 43800
    },
    {
      "epoch": 0.9150790011256097,
      "grad_norm": 1.2357829809188843,
      "learning_rate": 7.653211284818598e-06,
      "loss": 1.8403,
      "step": 43900
    },
    {
      "epoch": 0.9171634635427524,
      "grad_norm": 0.9447447657585144,
      "learning_rate": 7.336250059950618e-06,
      "loss": 1.8304,
      "step": 44000
    },
    {
      "epoch": 0.9192479259598949,
      "grad_norm": 1.0322580337524414,
      "learning_rate": 7.0221442185559375e-06,
      "loss": 1.8415,
      "step": 44100
    },
    {
      "epoch": 0.9213323883770376,
      "grad_norm": 0.8802585601806641,
      "learning_rate": 6.711230463646687e-06,
      "loss": 1.85,
      "step": 44200
    },
    {
      "epoch": 0.9234168507941802,
      "grad_norm": 1.3486437797546387,
      "learning_rate": 6.4038420765057915e-06,
      "loss": 1.8361,
      "step": 44300
    },
    {
      "epoch": 0.9255013132113228,
      "grad_norm": 1.2704123258590698,
      "learning_rate": 6.100308559428989e-06,
      "loss": 1.8299,
      "step": 44400
    },
    {
      "epoch": 0.9275857756284654,
      "grad_norm": 0.9107852578163147,
      "learning_rate": 5.800955282517735e-06,
      "loss": 1.833,
      "step": 44500
    },
    {
      "epoch": 0.929670238045608,
      "grad_norm": 0.9406431913375854,
      "learning_rate": 5.50610313490157e-06,
      "loss": 1.8598,
      "step": 44600
    },
    {
      "epoch": 0.9317547004627507,
      "grad_norm": 0.8901939392089844,
      "learning_rate": 5.2160681807638155e-06,
      "loss": 1.8485,
      "step": 44700
    },
    {
      "epoch": 0.9338391628798933,
      "grad_norm": 0.8982817530632019,
      "learning_rate": 4.931161320539333e-06,
      "loss": 1.8423,
      "step": 44800
    },
    {
      "epoch": 0.9359236252970359,
      "grad_norm": 1.0009591579437256,
      "learning_rate": 4.6516879576475195e-06,
      "loss": 1.8406,
      "step": 44900
    },
    {
      "epoch": 0.9380080877141785,
      "grad_norm": 1.0994175672531128,
      "learning_rate": 4.377947671117785e-06,
      "loss": 1.8281,
      "step": 45000
    },
    {
      "epoch": 0.9400925501313211,
      "grad_norm": 0.9065024852752686,
      "learning_rate": 4.110233894458426e-06,
      "loss": 1.8417,
      "step": 45100
    },
    {
      "epoch": 0.9421770125484638,
      "grad_norm": 0.9834316372871399,
      "learning_rate": 3.848833601113141e-06,
      "loss": 1.8518,
      "step": 45200
    },
    {
      "epoch": 0.9442614749656064,
      "grad_norm": 1.074263572692871,
      "learning_rate": 3.594026996842356e-06,
      "loss": 1.8217,
      "step": 45300
    },
    {
      "epoch": 0.946345937382749,
      "grad_norm": 1.2309187650680542,
      "learning_rate": 3.3460872193591042e-06,
      "loss": 1.8357,
      "step": 45400
    },
    {
      "epoch": 0.9484303997998916,
      "grad_norm": 1.2959349155426025,
      "learning_rate": 3.1052800455414456e-06,
      "loss": 1.8349,
      "step": 45500
    },
    {
      "epoch": 0.9505148622170342,
      "grad_norm": 0.8000039458274841,
      "learning_rate": 2.8718636065352667e-06,
      "loss": 1.8236,
      "step": 45600
    },
    {
      "epoch": 0.9525993246341768,
      "grad_norm": 1.6720008850097656,
      "learning_rate": 2.646088111052847e-06,
      "loss": 1.8372,
      "step": 45700
    },
    {
      "epoch": 0.9546837870513195,
      "grad_norm": 0.9012230038642883,
      "learning_rate": 2.4281955771638177e-06,
      "loss": 1.8259,
      "step": 45800
    },
    {
      "epoch": 0.9567682494684621,
      "grad_norm": 1.3057595491409302,
      "learning_rate": 2.218419572866011e-06,
      "loss": 1.853,
      "step": 45900
    },
    {
      "epoch": 0.9588527118856047,
      "grad_norm": 0.9006788730621338,
      "learning_rate": 2.0169849657142825e-06,
      "loss": 1.8372,
      "step": 46000
    },
    {
      "epoch": 0.9609371743027473,
      "grad_norm": 1.2864965200424194,
      "learning_rate": 1.8241076817757041e-06,
      "loss": 1.8409,
      "step": 46100
    },
    {
      "epoch": 0.9630216367198899,
      "grad_norm": 0.9104560613632202,
      "learning_rate": 1.6399944741694962e-06,
      "loss": 1.8518,
      "step": 46200
    },
    {
      "epoch": 0.9651060991370326,
      "grad_norm": 0.9562528729438782,
      "learning_rate": 1.4648427014398336e-06,
      "loss": 1.8604,
      "step": 46300
    },
    {
      "epoch": 0.9671905615541752,
      "grad_norm": 0.9484128952026367,
      "learning_rate": 1.2988401159990793e-06,
      "loss": 1.8423,
      "step": 46400
    },
    {
      "epoch": 0.9692750239713178,
      "grad_norm": 0.9366037845611572,
      "learning_rate": 1.1421646628682215e-06,
      "loss": 1.8204,
      "step": 46500
    },
    {
      "epoch": 0.9713594863884604,
      "grad_norm": 0.9006665349006653,
      "learning_rate": 9.949842889302675e-07,
      "loss": 1.8443,
      "step": 46600
    },
    {
      "epoch": 0.973443948805603,
      "grad_norm": 0.8474847078323364,
      "learning_rate": 8.574567629010489e-07,
      "loss": 1.8455,
      "step": 46700
    },
    {
      "epoch": 0.9755284112227457,
      "grad_norm": 0.9436659812927246,
      "learning_rate": 7.297295062104215e-07,
      "loss": 1.8402,
      "step": 46800
    },
    {
      "epoch": 0.9776128736398882,
      "grad_norm": 1.5559815168380737,
      "learning_rate": 6.119394349751628e-07,
      "loss": 1.8475,
      "step": 46900
    },
    {
      "epoch": 0.9796973360570309,
      "grad_norm": 1.059956431388855,
      "learning_rate": 5.042128132329205e-07,
      "loss": 1.8565,
      "step": 47000
    },
    {
      "epoch": 0.9817817984741735,
      "grad_norm": 1.0230424404144287,
      "learning_rate": 4.0666511759459457e-07,
      "loss": 1.845,
      "step": 47100
    },
    {
      "epoch": 0.9838662608913161,
      "grad_norm": 0.9298943281173706,
      "learning_rate": 3.19400913460185e-07,
      "loss": 1.8235,
      "step": 47200
    },
    {
      "epoch": 0.9859507233084588,
      "grad_norm": 1.1216191053390503,
      "learning_rate": 2.4251374293082885e-07,
      "loss": 1.8314,
      "step": 47300
    },
    {
      "epoch": 0.9880351857256013,
      "grad_norm": 1.3075478076934814,
      "learning_rate": 1.7608602453716205e-07,
      "loss": 1.8308,
      "step": 47400
    },
    {
      "epoch": 0.990119648142744,
      "grad_norm": 0.9057138562202454,
      "learning_rate": 1.2018896489149333e-07,
      "loss": 1.8327,
      "step": 47500
    },
    {
      "epoch": 0.9922041105598866,
      "grad_norm": 0.9391676783561707,
      "learning_rate": 7.488248235850038e-08,
      "loss": 1.8382,
      "step": 47600
    },
    {
      "epoch": 0.9942885729770292,
      "grad_norm": 0.8697664141654968,
      "learning_rate": 4.0215142826263155e-08,
      "loss": 1.8641,
      "step": 47700
    },
    {
      "epoch": 0.9963730353941719,
      "grad_norm": 1.663161277770996,
      "learning_rate": 1.6224107646479747e-08,
      "loss": 1.8362,
      "step": 47800
    },
    {
      "epoch": 0.9984574978113144,
      "grad_norm": 0.9440004229545593,
      "learning_rate": 2.935093799688593e-09,
      "loss": 1.8277,
      "step": 47900
    },
    {
      "epoch": 1.0,
      "step": 47974,
      "total_flos": 5.824755462670647e+18,
      "train_loss": 2.213516155517568,
      "train_runtime": 14249.5377,
      "train_samples_per_second": 26.934,
      "train_steps_per_second": 3.367
    }
  ],
  "logging_steps": 100,
  "max_steps": 47974,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": false,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.824755462670647e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}